Upper Layer Temperature - CFAST 7.2.4 and FDS 6.6.0 Comparison (Issue?)

[max.boe...@gmail.com]

Hello,

as part of a project, I compare the simulations results of FDS 6.6.0 and CFAST 7.2.4 in the calculation of the upper Layer Temperature during a full room fire. The basis for determining the heat release rate is the European Standard DIN EN 1991(Action on Structures).

The simulation setups for FDS and CFAST are identical. The room Dimensions are 15 m x 15 m x 3 m. There are a total of 50.7 m^2 of ventilation area to make sure that enough oxygen inflow is available. The fire has the following characteristics. Chemical composition is C58H93O11N1 with a Heat of Combustion of 30.000 kJ/kg. It is assumed, that the entire room (floor) is burning (assumption is given by DIN EN 1991).

The measurement of the ULT is done with four DEVC-Lines distributed in the room and by measuring the Quantity 'UPPER LAYER TEMPERATURE'. For the comparison with CFAST, there average of the four DEVC Lines are taken into account. CFAST is using the Normal (Two-Zone-Model) with adiabatic Compartment Surfaces (the last one is also correct for FDS).

Those characteristics are identical for each simulation.

A total of eight simulations were calculated. Each calculation with a different HRR (5 MW, 10 MW, 20 MW, 30 MW, 40 MW, 50 MW, 60 MW, 70 MW; also given by DIN EN 1991)

FDS and CFAST tend to give out increasingly differences in Temperature with higher HRR. In my opinion CFAST calculations are unrealistic high, as shown in the following graphic:

Are these results to be expected (because of nature of CFAST) or are they calculated incorrectly by CFAST?

I have attached the CFAST and FDS Input file and also HRR-Graphs for each Simulation.

Thanks in advance for your help.

Max Boehler. 

[Kevin McGrattan]

My guess is that you've set the walls to be adiabatic in the CFAST run, and CFAST does not compute radiative loss through the windows. So you've created a perfect furnace.

[max.boe...@gmail.com]

Hello Kevin,

thanks for your answer. Is there a way to simulate the radiative heat loss through windows/vents?

[Kevin McGrattan]

Hmmm, that's an interesting idea.

[andyta...@yahoo.com]

Dear all,

I am starting to initiate a minor change to the existing radiation solver to account for the radiation loss (fire, gas, and wall emission) due to openings for a compartment. 

This attempt is based on a ratio in between the area of openings associated to a wall segment and the total area of the wall segment. Based on my previous experience, the accuracy of this simplified approach is generally acceptable. I will also create a case and compared the results with radnnet-zm and investigate the potential differences.

I will come back to this post once I have some preliminary results.

P.S. I will be focusing on single compartments first. 

Thanks,

Andy

[max.boe...@gmail.com]

Hello Andy,

That sound interesting. I am looking forward to it!

Thanks,

Max

[andyta...@yahoo.com]

Dear Max,

I added a few lines of codes to account for the energy loss to openings from different components of radiation heat transfer and a quick calculation was done. The radiation loss associated to a wall segment is approximated based on the ratio of the total opening area on the wall segment to the total surface area of the wall segment.

For 70MW case, the upper layer temperature will be reduced to ~1250degC. And for 30MW case, the upper layer temperature will be reduced to ~670degC.

The preliminary results seem reasonably promising in your case.

Yet, there will still be a number of works that have to be done and I will have to seek advice from my colleagues.

Thanks,

Andy

[max.boe...@gmail.com]

Dear Andy,

sorry for my late answer. The results sound promising. As I see, you already pushed it to GitHub (#1149). I will try the code next week.

Thanks for your help!

Have a nice weekend,

Max.

[andyta...@yahoo.com]

Dear Max,

The implementation has not been fully completed yet.

The update only includes the calculation of the surface opening fraction associated to each surface.

The surface opening fraction is going to be used for the evaluation of convective heat exchange between gas and surfaces and the radiation heat transfer. The implementation of surface opening fraction to radiative heat exchange calculation will need careful considerations especially when we are dealing with multiple-compartments.

Thank you very much for your patience.

Kind regards,

Andy